US3503064A - A-d conversion system - Google Patents

Description

March 24, 1970 TAKEKI TAKARABE ETAL 3,503,064

A-D CONVERSION SYSTEM 3 Sheets-Sheer@ 1 Filed Dec. 21, 1964 V/WIIIIII:

March 24, 1970 TAKEKl TAKARABE ET L 3 Sheets-Sheet 5 Filed Dc. 21, 1964 ke'l, EVENTORS rabe 3h0 g/Na rafa/71 United States Patent O U.S. Cl. 340-347 1 Claim ABSTRACT oF THE DISCLOSURE An analog-to-digital conversion system which allows for the compensation of a non-linear detecting element so as to produce an accurate system.

This invention relates to an A-D conversion (analogto-digital conversion) system, and in particular to an A-D conversion system in which A-D conversion is carried out `while compensating the non-linearity of a detecting element in the course of conversion of analog signals from the detecting element into digital signals.

According to measuring systems heretofore employed, when analog signals from detector ora signal generator areconverted into digital signals a preamplifier is used so as,- to arrange at a certain level the signals from various detectors or signal generators, and when the output signals .of such signal generator do not vary linearly in response to voltagevor resistance variations caused due to a measured value such, for example, as temperature variations, a complicated circuit such as a secondary converter is provided to compensate the non-linearity. The signals, which have thus been made uniform, are then supplied to an A-D conversion circuit.

However, the conventional measuring systems have a disadvantage in that highly precise measurements cannot be performeddue to drift and oifset of the preamplier', variations in the characteristics of an input circuit time constant and linearity thereof with the lapse of time and precision of the non-linearity compensation of the secondary converter, other than errors of the A-D converter circuit.

Accordingly, one object of this invention is to provide an A-D conversion system in which D-A conversion (digital-to-analog conversion) of signals from a detector or a signal generator is directly carried out and operations including the non-linearity 4compensation are simply performed with high precision Another object of this invention is to provide a simple and effective A-D conversion system comprising a circuit portion for directly performing a D-A conversion of" signals delivered from a detecting element irrespective of resistance or voltage variations caused by the detecting elementitself and a circuit portion for compensating the non-linearity of the detecting element, in which during the D`A conversion the non-linearity is cornpensated and an A-D conversion of such compensated amount is carried out at the same time without necessitating the-use of a complicated secondary converter.

A further object of this invention is to provide an A-D conversion system in which a Wheatstone bridge circuit is formed With a detecting element or a signal generator, av groupv of D-A conversion measuring resistors, nonlinearity compensating resistors and a proportional resistor and D-A conversion of the detected value is carried out together with the non-linearity compensation.

Another object of this invention is to provide an A-D conversion system in which a detector or a signal generator is incorporated in one arm of a Wheatstone bridge circuit and its output side is connected to a D-A oonverter and the non-linear distortion of the detector or signal generator is compensated during the operation of the D-A converter, thereby performing accurate digital measurements.

Other objects, features and advantages of this invention will be apparent from the following description taken in conjunction with the accompanying drawings, in which:

FIGURE 1 is a connection diagram illustrating an example of the system of this invention, which comprises a resistance-type D-A converting device formed by incorporating a thermal bulb, a group of D-A conversion measuring resistors, non-linearity compensating resistors for the thermal bulb and a proportional resistor into respective arms of a Wheatstone bridge circuitl and a voltage-type D-A converting device having a thermo-electric couple as a detecting element, these D-A converting devices being changed over to carry out D-A conversion in accordance with their respective types;

FIGURE 2 is a graph for explaining a. temperature characteristic of the thermal bulb and compensation of its non-linearity;

FIGURE 3 is a graph of characteristic curves illustrating the relationship of variations in the level of some liquid tank to those in capacity thereof, and for explaining thatthe characteristic curve of the tank can be ap- Y proximated by three straight lines;

FIGURE 4 is a connection diagram illustrating an embodiment of this invention in which the capacity of the tank approximated by the three straight lines shown in FIGURE 3 is converted into a digital form.

The present invention can be practised in the resistance or voltage type in accordance with the kind of the detecting element or signal generator, and an example of this invention will hereinbelow be explained with reference to FIGURE 1 in which either one of the two types can be selected as desired.

In FIGURE l, 1 is a platinum thermal bulb used for, for example, temperature measurement and it has a resistance value of, for instance, 1009. The platinum thermal bulb 1 is inserted in one arm D of a Wheatstone bridge circuit W and a resistance-type D-A conversion circuit is formed which has arms A, B and C. That is, a power source 7 is connected between a pair of diagonal points t1 and t2 of the bridge circuit W tand a comparison amplifier 4 is connected between another pair of diagonal points t3 and t4 through lead wires 2 and 3 and a switch S3. Non-linearity compensating resistors for the thermal bulb 1 are incorporated into the arm A of the bridge circuit W, namely resistors a0, a1 and a2 of, for example, 1009 are connected in series and the resistors S1 and S2 respectively. A proportional resistor b1 of 10052 is incorporated into the arm B. In the arm C measuring resistors R1, R2, R4, Rs, R10, R20, R40, Rao, R100, R200, R400, R800 are Connected in series with a base resistor R0. The resistors R1 to R800 are respectively given in order of unit resistance values of 800, 400, 200, 80, 40, 20, l0, 8, 4, 2 and 1 based upon the binary scale and short-circuit contacts P1, P2, P4, Pa, P10, P20, P40, P80, P100, P200, P400 and P000 are PTO vided which respectively correspond to the resistors R1 to R800. In practice, resistance variation of the platinum 3 thermal bulb is generally .3869 per 1 C., and hence the resistance values of the aforementioned resistors R1 to R400 are respectively selected in accordance with the following table.

The resistors a1 and a2 are used for non-linearity compensation as will be explained later and n the present example they show gnadients at 200 C. and 400 C., and they have values of 65.759 and 118.269 respectively. A controller 5, for successively switching on and off the contacts P000 to P1 to thereby operate the contacts S2 and S1, and a timing circuit 6, for operating the controller 5, are provided. That is, the controller 5 consists of many relays L0000, L0400 L04, L02 and L01 comprising contacts P0000 P00, P02 and P01 which drive conversion relays of the respective units of a binary-decimal conversion circuit 7 in response to the contacts P000 to P1 and the resistors R000 to R1. Since the binary-decimal conversion circuit 7 is a well-known one, its detailed connection is not illustrated in the drawing for the sake of simplicity. To facilitate better understanding of the operation, the relays and the contacts are connected by the dotted line DL1.

The timing circuit `6 may be constituted in the form of the so-called ring counter, namely it includes relays Lrson, Lrzoo Lrzoo Lr4, Lrz and Lri 1eSPeetlVely COT' responding to the controller relays L0000, L0400, L0200 L04, L02 and L01. For operating the ring counter 6, a timing oscillator 8 and a flip-flop circuit 9 to be controlled thereby are provided. The timing oscillator may be constituted in the form of a multivibrator which produces a pulse of, for example, 1() c./s., the frequency of which pulse is reduced by half in the flip-flop circuit 9. On the output side of the timing oscillator 8 and the :flip-flop circuit 9, a pair of transistor switches 10 and 11 are provided, through which the respective relays L0-s of the ring counter I6 are operated in turn each for 0.1 second, thereby successively actuating the controller relays L0's respectively corresponding to the relays L00.

The relays L00 are provided with four contacts respectively and, for example, L0000 has contacts P0001, P0002, Praos and Prao4 and Lr4oo has Contacts Pr4o1, Puna, Puna and P0404. In the like manner, the remaining relays of this kind are each provided with four contacts. Since the relfys Lrsoo, Lr4oo, Lrzoo I'ellys Leann, Leioa, L0200 and their respective contacts are similarly connected to other parts, we will explan the relay 1.0000 and its contacts P0001, P0002, P0000 and P0004 alone.

That is, one end of the relay L0000 is connected through a contact Ps of a start relay Ls to one output side 11a of the transistor switch 11 and the other end is connected through the relay L0 to the minus side of a power source E1. The contact P0001 is inserted in parallel with the contact PS and one end of the relay L0400 is connected to the other output side 11b of the transistor switch 1:1 through the contact 1,0002 formed late-break-type. The other end of the relay L0400 is connected to the minus side of the power source E1. The relay L0200 is similarly connected to the output side 11a of the transistor switch 11 through the corresponding contact P0402 of the relay L0400 and to the minus side of the power source E1. The other relays L00 are connected in the same manner. One end of the relay L0000 is connected through the contact P0000 to one output side 10b of the transistor switch 10 and the other end is connected through a resistor R0000 to the minus side of a power source E2. One end of the resistor R0000 on the side of the relay L0000 is connected to the other output side 10a of the transistor switch 10` through the contact P0004 and a contact P01 described later. The other relays L0400, L0200 L02 and L01 and resistors R0000 to R01 are connected substantially in the same manner.

To the output side of the comparison amplier 4 is connected la transistor 12 and a relay L01 is connected to the ,output side thereof, by which it is discriminated whether the relays L0000 to L01 of the controller 5 must be held or reset. To perform this, the contact Pd is provided on the output side 10a of the transistor switch 10 as connected by the dotted line D112. That is, the relay, for instance L0000 is held or reset according as the voltage of the resistor R0000 drops or not at the time of make of the contact P0004 in response to make and break of the contact P0. E0 is a DC power source. It is preferred to connect diodes D in parallel to the'A aforementioned relays respectively so as to stabilize their operation.

The operation of the system of this invention fwill hereinafter be explained. When the switch S0 is changed over from the position -as illustrated in FIGURE 1 the terminals t0 and t4 Yof the Wheatstone bridge circuit W are connected to the input side 'of the comparison amplitier 4 and the switch S of the bridge circuit W is closed. At this time, the thermal bulb 1 is incorporated into the arm D of the bridge circuitW, which bulb Ivaries its resistance value in response to temperature of a portion to be measured. Then, the resistance values of the thermal bulb 1 and the arm C at this time compared. The contacts P000 to P1 corresponding to the resistors R000v to R1 areU normally closed.

When the contact P0 is closed by the start relay L0 the relay L0000 is energized to close its contacts P0001,"P00'02, P0000 and P0004. It must be noted that the relaysfL0000' 1.0400, 1,0200 are energized in turn each for 0.1 second through the transistor switches 11a and 11b'.0In this case, the contact P0001 remains closed for 0.1 second and immediately opens, but the contact P0002 is of Ilate-break type and still remains closed for a short period of time after the relay 1,0000 has been deenergized. Hence, the relay 1.0400 is similarly energized for 0.1 second through the contact P0002, thereby closing the contacts P0401, P0402, P0400 and P0404, closing the contact P0000, a current is applied to the relay L0000 and the resistor R0000 by the output of the transistor 10'b. By the energization of the relay L0000 the contact P000 is opened. At this time, the bridge circuit W carries out measurements and applies the results to the comparison amplifier 4, in which the resistance values of the thermal bulb 1 and the resistor R000 are compared. In case the resistance value of the resistor R000 is greater than that of the thermal bulb 1 in this comparison, the relay L01 of the discriminator is conducted to close its contact P01. As a result of this, one end of the resistor R0000 passing through the contact P0004 is grounded to decrease the quantity of conduction to the relay L0000, holding the relay L0000 inoperative and closing again its contact P000. Simultaneously with it, the contact P0000 opens to keep the binary-decimal converter circuit 7 out of operation. Then, the relay 1.0400 carries out similar operation and comparison. Also in this case, when the resistance value of the resistor R400 exceedsv that of the thermal bulb 1 the contact P400 closes and the contact-0 P0400 opens. However, when the resistance value 0f the resistor, for example R200 is less than that of the thermal4 the contact P0200 closes and the binary-decimal converter circuit 7 operates.

Thus, the relays L0-000 to L01 operate one Aafter another each for, for example, 0.1 second and the controller` relays L0000 to L01 also operate each for 0.1 second correspondingly. This operation is performed repeatedly. Accordingly, when any one of the resistance values of the resistors R000 to R1 exceeds that of the thermal bulb 1, its corresponding relay Lc becomes inoperative to close again its corresponding contact which has once opened. The relays Lys succeeding the above-mentioned one, accordingly their corresponding relays Le', are held closed. On the contrary, when the resistance value of any one of the resistors Rs is less than that of the thermal bulb 1 its corresponding contact opens and its corresponding resistor is connected to the bridge circuit W. Consequently, the sum of the resistance values of those which have been opened by a series of such operation of the ring counter among the contacts P000 to P1, equals the resistance value of the thermal resistor 1 and the binarydecimal converter circuit 7 is operated through contacts of those which have opened among the relays L05, thereby indicating the temperature in a numerical value 'with an indicator 13.

Where temperature, which is an amount to be measured, is measured by using the thennal bulb in the temperature measuring circuit described above, the temperature characteristic of the thermal bulb is as shown by the curve y in FIGURE 2 (the abscis-sa expressing temperature T and the ordinate response Y), and the resistance value of the thermal bulb 1 varies in a non-linear manner with respect to temperature. However, it can be expressed by asymptotes divided as desired in accordance rwith precision required. It can approximately be expressed by the straight lines y1 and y2 along the curve y having a turning point P atv the point of a temperature 150 C., for example. That is,

y2=a2x+b2 Where values of a1 and a2 are gradients of the curves y1 and y2. To determine them, it is suicient only to change the values of a1 and a2 of the proportional arm A of the bridge circuit W. The values of b1 and b2 may be obtained by correcting or changing the D-A converting resistors on the arm C of the bridge circuit W. R0 shows a base value that, for example, the straight line y2 crosses the ordinate and it has been selected 100.3 in this example. That is, since C B D in the bridge circuit W shown in FIGURE 1, a1 and a2 are determined by the value of A/B and in practice the values of the arm A are changed over successively in the process of comparison. In the present example a platinum thermal bulb of 1000 is employed, so that a resistor of 1009 is inserted into the arm B. Furthermore, for example, 150" C. is a reference and the straight lines y1 and y2 are considered to cross each other at this point. Therefore, the -value of the resistor a1 is selected to be 65.7512, which corresponds to 200 C., and its short-circuit contact S1 is so designed as to operate together with the contact P200. Similarly the value of the resistor a2 is selected to be 118.269, which corresponds to 400 C., and its short-circuit contact S2 is also formed to operate together with the contact P400. If now a temperature to be measured is 250 C., the contact P400 of the relay Rc400 and the contact S2 open in the aforementioned comparing operation but the resistors R400 and a2 are too great for 250 C., and hence the contacts P400 and S2 close immediately. Then, the contacts P200 and S1 open when the relay R0200 operates. In this case, since the resistance value of the resistors R200 and a2 is smaller than that corresponding to 250 C., the relay Rc200 remains held and the contacts P200 and S1 remain open, thus carrying out the nonlinearity compensation of the temperature.

We will explain an example in which the present invention has been applied to the voltage type measurement using a thermoelectric couple which produces temperature variations directly in the form of voltage. That is, 14 shows a thermoelectric couple and 15 its cold junction compensator. In this example, terminals of both ends of a DC power source 70 are designated at t5 andtq and a terminal t0 is connected through a resistor R0' to the terminal t1. Between the terminals t5 and t6, there are connected in parallel D-A conversion circuit resistors R200 to R1 respectively corresponding to the aforementioned resistors R000 to R1, and contacts P000 to P1 are provided in series respectively with these resistors. These contacts correspond to the above-mentioned contacts P000 to P1 and perform the same operation. They are controlled by the relays L0200 to L01 of the controller 5 as illustrated by the dotted lines D12. The value of the resistors is determined in accordance with that of the resistor R0 and the variation ratio of the voltage of a detector. Where the resistor R0 has a value of 3.0709 and a CA -thermoelectric couple is used, numerical values of the resistors R000 to R1', by way of example, are as given in the following table.

The thermoelectric couple 14 is connected between the terminals t2 and t0 and the cold junction compensator 15 is connected between terminals t0 and t0 in series with the thermoelectric couple 14. Thus, a circuit is formed which compares a voltage between the terminals of the resistor R0 and a voltage produced in the thermoelectric couple '14, and a difference voltage therebetween is impressed to the input side of the comparison amplifier 4. Furthermore, non-linearity compensating resistors a1 and a2', which are similar to those mentioned in the foregoing, are connected between the terminals t0 and t7 in parallel with the base resistor R0'. Also in the voltage-type measurement, there is the non-linearity in the thermoelectric couple and its value varies with measured temperatures like in the aforementioned resistance-type measurement. To compensate the measured voltage of the D-A conversion circuit, a resistance value corresponding to R0 is changed. To perform this, the resistors a1 and a2' are incorporated into the circuit by means of contacts S1 and S2' respectively and their values represent gradients of straight lines which have approximated the characteristic curve of the thermoelectric couple. In the present example the -resistors a1 and a2 are selected to be 75 .470 and 12l.4l respectively. The contacts S1 and S2', inserted in series to the resistors a1' and a2 respectively, are so formed as to be operated by, for example, the relays L0400 and 1.0800 respectively. Also in the voltage-type measurement a term b is considered with respect to the formula previously mentioned as in the resistance-type measurement, but this can be obtained by compensating the resistance value at a predetermined position of the resistors R1 to R000. l

The terminals t0 and t9 are connected to the input side of the comparison amplifier 4 and the switch S3 is employed to select either the resistance-type measuring system or the voltage-type one as desired.

Also in the voltage-type measurement above described, the contacts P000 to P1' and S1 to S2 are successively changed over to carry out the comparing action and temperature is measured by the use of the thermoelectric couple, while compensating the non-linearity due to the thermoelectric couple.

Since similar operations are effected for both the voltage-type and resistance-type measurements as has been 2 FIGURE 1 shows the resistance-type and voltage-type measuring circuits together, illustrating an example in which the two circuits are changed over by the switch S and the comparison amplier y4, the discriminator L01, the controller 5, the timing circuit `6 and so on are used in common. However, either the resistance-type measurement or the voltage-type one can be independently used as required. In short, it will be seen that the present inventioncan be applied to both the resistance-type and voltage type measurements.

The foregoing has been made in connection with an example in which a non-linear line is approximated by two straight lines passing through a turning point. Then, we will explain a case where a non-linear line can be approximated by three asymptotes passing through two or more turning points. A liquid storage tann of, for 'example, ships is required to store liquid as much as possible in the smallest possible space, and hence the tank-is never simple in shape and usually takes a specific shape in accordance with conditions of its location. This is an example in which a capacity value of such tank is computed from a measured level. FIGURE 3 illustrates a characteristic curve of a tank, the abscissa expressing indicating values of the tank capaicity in m.3 and the ordinate the liquid level of the tank in meter and the output voltage V of a leveler corresponding thereto in volts. In such a case the tank shows an actual tank capacity curve such as shown by the full line y.

This full-lined curve can be approximated by a plurality of dotted-lined asymptotes y1, y2 and ya drawn along the full-lined curve. That is,

Where b1, b2 and b5 respectively show distances between intersecting points of the straight lines with the ordinate and the origin `O. Computation of a formula expressing such straight lines can be concretely accomplished by a circuit shown in FIGURE 4. In the ligure R000l to R1" are resistors connected in parallel between terminals t15 and t15 and they are respectively given resistance values based upon the binary scale of 800, 400 4, 2 and 1. These resistors are switched on and off by contacts P000" to P1" connected in series thereto respectively and they are controlled by a controller according to their respective positions as in the foregoing example. The terminals of a power source 70 are designated at t15 and r11, and a base resistor R01 is connected between the terminals t1@ and IT.

OS is an oscillator which produces voltage V corresponding to the liquid level of the tank, namely to the level and the oscillator is represented by a DC power source 71 and a variable resistor R02 connected thereto. a is a portion for determining the terms of a1, a2 and a5 in the aforementioned formula and it comprises resistors R01, R02 and R02 and contacts S10, S11 and S12 respectively connected in series to the resistors through a switch S4 and a resistor R05 connected in series to the output terminal t of the oscillator OS, the resistors being connected in parallel to each other between terminals t10 and 110. At the next stage of the portion a, namely between ythe portion a and a terminal t11 a portion b is provided for determining b1, b2 and b5 in the foregoing formula. The portion b comprises resistors R05 and R04 connected between the portion a and the terminal r11, a power source 73 connected to the connecting point of the resistors, switches S5 and S0 connected to the both ends of the power source, a resistor P01 and a switch S1 respectively connected in series to the outer ends of the switch S5 and the resistor R03, parallel resistors R02 and P05 connected between the outer ends of the switch S5 and the resistor R01 switches S0 and S5 respectively connected in series thereto. The switches 8.1, S5 and S11-are so formed asto operate together at the time of changing over the input.

The switches S1 to S12, inclusive, carry out make and break action in accordance with a sequence predetermined in the process of the binary action of thecontroller 5.

In the above example the values of the resistors R01 and R000 to R1", inclusive,` are respectively constant as shown in the following table and the input signal coming from the oscillator OS is approximated by straight lines lin the course of the A-D conversion and a voltage across the resistor R01 and an input signal voltage between the terminals r11 and t10 are compared. At this time the input signal is compensated. Of the three straight lines, y1 covers a range from 0 to 9.9 cm.3 and its gradient a1 is determined by the resistance values of the resistors R05 and R01 and further the distance b1 from the origin O is determined by a Voltage across the resistor R01. Since the resistors Ral and R04 respectively produce voltages, the contacts S5 and S0 close as shown by the dotted line. In such a case, the contacts S5 and S0 have already closed as above described, and hence a current passing through the resistor R04 is the same in direction and in polarity as a current passing through the resistor R01.

In like manner y2 covers a range from 10` to 29.9 m.3v

and its gradient a2 depends on the value of the resistor R02 and further b2 is determined in accordance with a voltage of the resistor R01 due to insertion of the resistor R02, caused by the closing of contact S0. Its polarity is in the same direction as the signal, namely positive. Furthermore, y2 covers a range of more than 30 m and its gradient a3 is determined by the value of the resistor R03 due to the closing of the contact S12. The distance b5 depends upon a voltage of the resistor R03 due to insertion of the resistor P01 caused by the closing of the contact S0. The conducting direction of the resistor lR03 is opposite to that ot' the signal and this implies that it is inverse in polarity, namely it is in the position of b3 below the origin.

The respective resistors forming the circuits a and b are given suitable values respectively in such a manner that the capacity value of the tank may be indicated in digital form when the voltage of the power source 70 is 10 volts, the value of the resistor.R01 is 10Q, the resistors R000 to R1 are given predetermined values and the capacity of the tank is approximated by the three straight lines y1, y2 and y5 correspondingly as above described. An example of their numerical values is given in the following table.

Some of the above resistors are provided with their respectively corresponding contacts or switches and these contacts or switches are respectively controlled by the relays R00 operated in accordance with the respective resistance values.

Since the above-described circuit portions for actuating the contacts in a predetermined order in this example are the same as those explained in connection with FIGURE 1, portions corresponding to those in FIGURE 1 are marked with the same numeral references.

With an arrangement described in the foregoing, the resistors R000" to R1 of the D-A conversion circuit for the binary operation are actuated in turn from higher units to lower ones by the A-D conversion operating signal from the timing circuit 6. Then, a terminal voltage produced in the resistor R01 and an output voltage signal from the oscillator OS, namely a voltage produced be` tween the terminals tlf, and tlg' are successively compared in response to the closing of the respective resistors, thus computing digital amounts. At this time the non-linearity compensation is carried out on the side of the signal simultaneously as described previously.

It will be apparent that many modifications and variations may be eiected without departing from the scope of the novel concept of this invention.

What is claimed is:

1. An analog-digital conversion system comprising a detecting element for converting a monitored function into a voltage, a rst resistor in series with a first terminal of said detecting element, a comparison circuit with one terminal connected to a second terminal of said detecting element, a rst plurality of resistors and a rst plurality of switches connected in series with said rst plurality of resistors and the combination connected in parallel between the second side of the rst resistor and the second terminal of said detecting element, a second Iplurality of resistors and a second plurality of switches in serieS with said second plurality of resistors and the combination connected to the second side of the rst resistor and said second plurality of switches and said second plurality of resistors forming a second circuit, a power source connected to said second circuit, a second resistor connected in series between said second circuit and the one terminal of the comparison circuit, a third plurality of resistors and a third plurality of switches connected in series with the third plurality of resistors and the combination connected in parallel, a second power source connected to said third plurality of resistors, said third plurality of resistors connected to said comparison circuit, a discriminator connected to said comparison circuit, a controller connected to said discriminator and having relays which are connected to said first, second and third plurality of switches, a timing circuit connected to said controller, and an indicator connected to said controller to indicate the binary output of said detecting element.

References Cited UNITED STATES PATENTS 3,237,186 2/1966 Whelpley et al. 340-347 3,239,833 3/1966 Gray 340-347 3,245,072 4/1966 Fuller 340-347 3,255,449 6/1966 Euler 340-347 3,274,586 9/ 1966 Lapinski 340-347 3,283,319 11/1966 Kaneko 340-347 3,316,751 5/1967 Burk 340-347 MAYNARD R. WILBUR, Primary Examiner CHARLES D. MILLER, Assistant Examiner U.S. C1. X.R. 324-98 UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3,503,064 Dated March 24, 1970 Inventods) TAKEKI TAKARABE ET AL.

It is certified that error appears n the above-identified patent and that said Letters Patent are hereby corrected as shown below:

Column 2, line 57, after "resistors" should read --al and a2 are provided with short-circuit contacts.

Column "l, line 16, "tann" should read ta.nk.

Signed and sealed this 15th day of June 1971.

(SEAL) Attest:

EDWARD M.FIETCHER,JR. WILLIAM E. SCHUYLER, JR. Attesting Officer Commissioner of Patents FORM po'wso (w'sg) uscoMM-oc scan-Peo

Images